IEA Discussion Paper No.70

NEVER MIND THE GAP Why we shouldn’t worry about inequality

Ryan Bourne and Christopher SnowdonMay 2016

Institute ofEconomic A�airs

With some exceptions, such as with the publication of lectures, IEA Discussion Papers are blind peer-reviewed by at least one academic or researcher who is an expert in the field. As with all IEA publications, the views expressed in IEA Discussion Papers are those of the author and not those of the Institute (which has no corporate view), its managing trustees, Academic Advisory Council or senior staff.

3

Contents

About the authors 04

Summary 06

Introduction 08

Income inequality: UK trends 09

Income inequality: international comparisons 17

Wealth inequality: UK trends 20

Wealth inequality: international comparisons 25

Global income and wealth inequality 28

How should we think about inequality statistics? 31

Can inequality itself be the cause of problems? 35

Conclusion 41

References 44

Appendix: Piketty’s thesis and criticism 48

4

About the authors

4

5

5

Ryan Bourne is Head of Public Policy at the IEA and a weekly columnist for City AM. He has previously worked both at the Centre for Policy Studies and Frontier Economics and has written widely on a range of economic issues.

Christopher Snowdon is the Director of Lifestyle Economics at the IEA. He is the author of Selfishness, Greed and Capitalism, The Art of Suppression, The Spirit Level Delusion and Velvet Glove; Iron Fist. He has authored a number of IEA publications including Closing Time, Sock Puppets, The Proof of the Pudding, The Crack Cocaine of Gambling? and Alcohol and the Public Purse.

6

Summary

● After rising in the mid to late 1980s, most measures of income inequality peaked in around 1990 and have fallen since. The Gini coefficient, for example, is currently lower than it was in 1987 and is close to the EU average.

● Wealth inequality has not been rising rapidly in the past decade, and is actually lower than in most other developed countries. There remains a debate about the true level of wealth inequality in the UK, but the trends do not conform to the story of unprecedented or spiralling inequality that are frequently implied in the media. Wealth inequality recently widened for the first time in a decade. However, this was primarily due to housing wealth.

● A strong focus on inequality within nations obscures the fact that global income inequality has been falling. Net wealth inequality figures at a global level give a misleading picture of poverty and inequality. They fail to account for the demographic composition of the global population and imply, for example, that a rich westerner with large debts but few assets would find themselves at the bottom of the global distribution.

● The overwhelming focus on summary statistics of income or wealth at a given point in time perpetuates two misconceptions – namely, that a distribution can be easily controlled and that the economy is a zero-sum game.

● In reality, inequality statistics reflect the results of millions of individual interactions, exchanges, endowments and policies. Affecting a distribution of income or wealth inevitably means interfering with some of these trades or interactions. Saying you want to ‘reduce inequality’, absent conditionality of the causes, implies being willing to accept more poverty or less wealth overall provided the distribution is more narrow.

7

● Whilst inequality per se might not concern us, some causes of unequal outcomes which we might consider to be ‘unjust’ may well do. For example, the elimination of cronyism might be expected to improve general prosperity whilst also reducing the income gap.

● One might also be concerned about inequality if it could be shown that a larger gap between rich and poor led to a range of social problems, corrupted the democratic process or slowed economic growth. But the evidence for each of these mechanisms, constantly heard in the media, is weak.

● The big risk of this continued focus on inequality is that we lose sight of the goal of improving living standards generally and for the poorest in particular. Many, particularly on the political left, seem to be more interested in the riches of the rich than the poverty of the poor – or else imply that the first is the cause of the second.

8

Introduction

In 2013, Barack Obama described income inequality as ‘the defining challenge of our time’. A few months later, the Pope tweeted ‘Inequality is the root of social evil’. Ed Miliband, in a speech he gave as leader of the Labour party in 2014, said that tackling inequality was the ‘the new centre ground of politics’. He may have been right. David Cameron is quick to remind people that income inequality has declined under his stewardship. The 2015 Conservative Party manifesto boasted that ‘Inequality is down from the record high it reached under Labour’.

The Conservatives should be more cautious about taking credit for this supposed success story. As this discussion paper shows, inequality often declines when economies are performing badly. In any case, the post-war peak in inequality did not occur during Labour’s term in office but in 1990 under a Conservative government. On most measures, inequality has not increased since Gazza cried at the Italia ’90 World Cup and Bombalurina had a UK number one single with “Itsy Bitsy Teeny Weeny Yellow Polka Dot Bikini”.

Passions are stirred and misconceptions abound whenever inequality is discussed. Establishing the basic facts is a task in itself. Is it rising or falling? Globally or nationally? Are we talking about income inequality or wealth inequality? And that’s before we try to answer the important question: do these statistics even matter? This discussion paper provides the facts about economic inequality in the UK and offers some answers to that question.

9

Income inequality: UK trends

It is often claimed that income inequality in Britain is rising, but the data simply do not support this. The truth is that income inequality, as measured by the standard measure of the Gini coefficient, based on disposable household income, has not risen for a quarter of a century. As Figure 1 (taken from the Office for National Statistics) shows, there is ample evidence that inequality increased between 1977 and 1990, with most of the rise taking place in the mid to and late 1980s. Since then it has fluctuated without ever returning to its 1990 peak. The general trend has been downwards. By 2012, it was as low as it had been since 1986 (ONS 2016).

Figure 1: Income inequality in the UK (1977-2014/15) Gini coefficient

10

The reasons for the rise in income inequality in the 1980s are many and varied. In the early part of the decade, the British government took the decision to tie welfare payments to the rate of inflation rather than to wages. As a result, those who relied on benefits did not see their incomes rise as quickly as those who were in full-time work. When the economy was performing strongly and there was strong wage growth - particularly at the top - the gap between workers’ wages and benefit claimants’ incomes grew. Moreover, high rates of unemployment resulted in an increase in the number of claimants.

The flip side to this was that the unemployed, disabled and retired were protected from the effects of wage cuts in subsequent recessions. This is one reason why inequality fell during the recession of the early 1990s and after the financial collapse of 2008. In both instances, salaries declined in real terms - especially at the top - whereas benefits continued to rise in line with inflation.

Under Margaret Thatcher, the top rate of income tax fell from 83 per cent to 60 per cent and then to 40 per cent. Although these changes left more money in the pockets of higher earners, their role in exacerbating inequality can be overstated. The basic rate of income tax was also lowered (from 33 per cent to 25 per cent) and the top rate was not lowered until 1988, by which time inequality was near its peak. Thatcher’s tax-cutting policies arguably had a greater impact on inequality in a less direct fashion by attracting wealthy people to Britain and reviving the economy. The list of millionaires who fled Britain as tax exiles in the 1970s is long. Among the better known names are David Bowie, Michael Caine, the Rolling Stones, Rod Stewart, Sean Connery, Roger Moore and most of the Beatles. Those who remained, such as Elton John, ‘did so at enormous financial cost’ (Sandbrook 2012: 100). Driving high earners out of the country is one way to reduce income inequality but it benefits no one. This exodus began to be reversed in the 1980s and London, in particular, is now a desirable place to live for the global super rich.

The rise in income inequality was not unique to Britain. Around the world, technological progress disproportionately benefited highly skilled workers and globalisation opened up markets for a few players to become extremely wealthy. There was a rise in the number of pensioners (who tend to be poorer) and fierce competition for top executives which led to huge pay rises for many CEOs. Trade liberalisation across the globe is also likely associated with greater inequality in developed countries.

11

Given these trends have continued since 1990 it is perhaps surprising that inequality has fallen since. Policies such as the minimum wage have played a part at the low end of the distribution but the main explanation lies in Britain’s tax and benefits system which redistributes wealth on a grand scale. According to the Office for National Statistics, cash benefits reduced income inequality by 14.2 percentage points in 2014/15 while direct taxes reduced it by a further 3.2 percentage points. This progressive effect is only partially offset by regressive indirect taxes, such as VAT and tobacco duty, which increased inequality by 3.5 percentage points (ONS 2016).

In 2013/14, the household incomes of the richest fifth of Britons was £80,800 whereas the incomes of the poorest fifth was just £5,500. This is a 15:1 ratio. However, after taxes were taken and benefits given, this ratio fell to 4:1 (£60,000 and £15,500 respectively) (ONS 2015a). This is almost exactly the same ratio as in 1987.

The same picture emerges if we compare the incomes of the top and bottom fifth, or the 10th and 90th percentiles (shown in Figure 2), or the 90th and 50th percentiles.1 In every case, there was a significant rise in inequality between 1984 and 1990 followed by a stuttering, gradual decline.

1 The 20th percentile earns less than 80 per cent of the population but more than 19 per cent of the population. The 80th percentile earns more than 79 per cent of the population but less than the top 20 per cent.

12

Figure 2: Income inequality in the UK (1977-2014/15) P90/P10

There is one exception, however. If we look only at the proportion of income earned by the top one per cent it appears that the very highest earners have taken an increasing proportion of the nation’s pre-tax income, rising from around ten per cent in 1990 to fifteen per cent in 2010 (Piketty and Saez, 2012). Much of this change has been driven not just by the top one per cent, but by the top 0.1 per cent. Caution should be exercised when dealing with data from such a small number of households, but according to the World Top Incomes Database the share of gross income going to the top 0.1 per cent doubled from three to six per cent in the UK between 1993 and 2007/08 although this fell to 4.6 per cent during the economic downturn (ONS 2016).

The gap between the relatively rich (the 90th percentile) and the relatively poor (the 10th percentile) has declined since 1990, as Figure 2 shows. By contrast, the gap between the 90th and 99th percentiles has grown. The heated debate about ‘rising inequality’ is therefore not really about the difference between what a doctor earns and what a cleaner earns. It is about the difference between what a top lawyer earns and what a top footballer earns. It is about the rich versus the super-rich.

13

How have the top one per cent - and, in particular, the top 0.01 per cent - run away from the rest? Technology and globalisation have been particularly important in this respect by helping to create ‘winner takes all’ markets. Take Jeff Bezos, for example. Forty years ago, an entrepreneur of his kind might have had a bookshop in Seattle. If he had prospered, he might have had a chain of bookshops in the State of Washington. He might even have gone nationwide. Instead, thanks to the internet, he owns Amazon and can trade all over the world selling not just books but a vast range of products.

The internet has created the potential for vast audience magnification. Satellite television has turned British footballers into global superstars. Largely thanks to technology, barriers to entry are low and competition is fierce but the rewards are great for those who come out top in a global marketplace. To some extent, it is no more than a numbers game. The world’s population has increased from four billion to seven billion in the last forty years. There are simply more potential customers on the planet and they are wealthier and better connected than ever before. This is particularly true of markets with network effects, such as iPhone app. Producing an app in communication, such as Whatsapp, can lead to huge returns for the developer. This phenomenon – of being able to capture through entrepreneurial activity a huge global market – has been dubbed by Chicago economist Luigi Zingales as ‘winner takes all’.

The rise of the top one per cent may therefore be a cause of angst amongst the upper echelons of society. But it is not clear why it should concern the man on the Clapham omnibus. If low and middle earners were being dispossessed by the super-rich, it would be troubling but that is not what is happening. Despite the Great Recession, the average income in Britain was twice as high in 2013/14 as it was in 1977. Incomes in the bottom fifth rose more slowly, but still increased by 77 per cent in real terms (ONS 2015a: 12). Of course, where people make riches off the back of providing goods and services which enhance general utility, capturing a small chunk of that value for themselves, consumers also benefit in ways which may not always show up in measured income statistics.

In absolute terms, everybody has been getting richer. In relative terms, as Table 1 shows, the proportion of income earned by each decile in Britain has remained essentially unchanged for a generation. The top decile increased its share of income from 22 per cent in 1977 to a peak of 28 per cent in 1990. In 1987, the bottom decile saw its share fall from four

14

per cent to three per cent. Since then, however, there has been virtually no change in the share of any decile, rich, poor or average, whereas the absolute value of each share has increased significantly. The earnings of the bottom decile rose by 47 per cent (in real terms) between 1986 and 2011 (ONS 2012: 2). Three per cent of UK income today is worth a lot more than four per cent of income forty years ago.

15

Table 1: Percentage shares of equivalised disposable income by decile groups of all households, 1977-2012/13

Bottom 2nd 3rd 4th 5th 6th 7th 8th 9th Top

1977 4 5 6 7 8 9 11 12 15 221978 4 5 6 7 8 10 11 12 15 211979 4 5 6 7 8 10 11 13 15 211980 4 5 6 7 8 9 11 13 15 221981 4 5 6 7 8 9 11 12 15 231982 4 5 6 7 8 9 11 12 15 231983 4 5 6 7 8 9 11 12 15 231984 4 5 6 7 8 9 11 12 15 231985 4 5 6 7 8 9 11 13 15 231986 4 5 6 7 8 9 10 12 15 251987 3 5 5 6 8 9 10 12 15 261988 3 4 5 6 8 9 10 12 16 271989 3 4 5 6 8 9 11 13 16 261990 3 4 5 6 7 9 11 13 16 281991 3 4 5 6 7 9 11 13 16 271992 3 4 5 6 8 9 11 13 16 261993 3 4 5 6 7 9 10 12 16 27

1994-95 3 4 5 6 7 9 11 12 15 261995-96 3 5 5 6 8 9 11 13 15 251996-97 3 4 5 6 8 9 10 12 15 261997-98 3 4 5 6 8 9 10 13 16 261998-99 3 4 5 6 7 9 10 12 15 27

1999-2000 3 4 5 6 7 9 10 12 16 272000-01 3 4 5 6 8 9 10 12 15 272001-02 3 4 5 6 7 9 10 12 15 282002-03 3 5 5 6 8 9 10 12 16 262003-04 3 5 6 7 8 9 10 12 15 272004-05 3 5 6 7 8 9 10 12 15 262005-06 3 5 6 6 8 9 10 12 15 262006-07 3 4 5 6 8 9 10 12 15 272007-08 3 4 6 7 8 9 10 12 15 262008-09 3 5 5 6 8 9 10 12 15 262009-10 3 5 6 7 8 9 10 12 15 262010-11 3 5 6 7 8 9 10 12 15 272011-12 3 5 6 7 8 9 10 12 15 262012-13 3 5 6 7 8 9 10 12 15 26

Source: The Effects of Taxes and Benefits on Household Income, Office for National Statistics

1

Source: The Effects of Taxes and Benefits on Household Income, Office of National Statistics

16

When looking at these numbers, it is important to distinguish statistical categories from living, breathing human beings. People are not deciles. People move up and down this ladder in the course of their lives. Most of the people born into the bottom decile in 1977 will have been in a higher decile five, ten or twenty years later, and many of those in the higher deciles in 1977 will be in lower decile today. Claims about rising inequality are often made in tandem with claims about declining social mobility. Neither claim is true (see Snowdon (2015: 168-84) for the evidence on social mobility).

The point of showing this table is not to debate whether the distribution of income to any of the deciles is ‘fair’, but to look at what has happened to the richest decile since 1990. In short, nothing. Its share of income dropped to 26 per cent in 1992 and has stayed there or thereabouts ever since. A small subsection of the top decile may be increasing their share of income but this is not coming at the expense of the other 90 per cent. By any normal measure, inequality is not rising and has not risen for a very long time.

When the financial crisis hit in 2008 it was fashionable to predict that inequality would increase during the recession. It didn’t. From 2010, many people claimed that inequality would increase as a result of ‘austerity’. It didn’t. Conventional wisdom now says that further austerity cuts will lead to greater inequality. Perhaps they will, but this is far from certain since we do not know what the effect will be on employment and wages. The new National Living Wage may reduce inequality by raising incomes or it might increase it by causing unemployment. Much depends on how employers and individuals respond to changing incentives. What happens in the future is anyone’s guess, but there is no excuse for being unclear about the recent past and the present.

17

Income inequality: international comparisons

In addition to the claim that inequality is spiralling out of control, it is sometimes said that levels of inequality in Britain are exceptionally high. In recent years Britain has been variously described as ‘the fourth most unequal country in the world’ (Hassan 2012), the ‘European capital of inequality’ (Russia Today 2015) and ‘on course to take the no 1 spot for most unequal country in the developed world’ (Oxfam 2014). This is extremely misleading. Britain has traditionally had more income inequality than most of Scandinavia and Northern Europe, but it has never been Europe’s least ‘equal’ nation and is nowhere near the bottom of the world rankings.

The UK has unexceptional levels of inequality despite large disparities in salaries as a result of taxes and benefits, as the Office for National Statistics (2016) notes:

‘Before any taxes and benefits, the UK had one of the highest levels of income inequality in the EU. However, the UK’s tax and benefits system appears to be more redistributive than that of many other countries with relatively high pre-tax and benefits inequality, bringing the UK close to the overall EU average for inequality of disposable income.’

Looking at countries which the UN defines as being very highly developed, the UK has an average level of income inequality, according to the latest Gini figures from the World Bank (see Figure 3). Note that the World Bank does not hold inequality data for the richest Middle Eastern countries, such as Bahrain, Kuwait and Brunei, which are thought to have very high levels of inequality.

18

Figure 3: Income inequality in rich countries (Gini)

Table 1

Slovenia 25.59Norway 25.9Slovakia 26.12

Czech Republic 26.13Iceland 26.94Finland 27.12

Sweden 27.32Belgium 27.59Netherlands 27.99

Denmark 29.08Germany 30.13Austria 30.48

Hungary 30.55Switzerland 31.64Japan 32.11

Poland 32.39Ireland 32.52

United Kingdom 32.57

France 33.1Estonia 33.15

Montenegro 33.19Canada 33.68Cyprus 34.31

Luxembourg 34.79Australia 34.94Lithuania 35.15

Italy 35.16Latvia 35.48Spain 35.89

Portugal 36.04Greece 36.68United States 41.06

Argentina 42.28Israel 42.78Chile 50.45

New Zealand

Andorra

Bahrain

Brunei

Croatia

Hong Kong

Kuwait

Liechtenstein

Malta

Qatar

Saudi Arabia

Singapore

South Korea

United Arab Emirates

SloveniaNorway

SlovakiaCzech Republic

IcelandFinland

SwedenBelgium

NetherlandsDenmarkGermany

AustriaHungary

SwitzerlandJapan

PolandIreland

United KingdomFrance

EstoniaMontenegro

CanadaCyprus

LuxembourgAustraliaLithuania

ItalyLatviaSpain

PortugalGreece

United StatesArgentina

IsraelChile

0 15 30 45 60

1

19

The UK has relatively high levels of income inequality compared to OECD countries and much of Northern Europe, but it is average by the standards of the wider rich world. By global standards, Britain’s gap between rich and poor is also neither large nor small. The least equal nations are mostly in Central and South America where the Gini coefficient frequently exceeds 50, as compared with 32 in the UK and 26 in the rich world’s most equal country, Slovenia. Some ex-Soviet states have lower levels of income inequality than Britain but nearly every African country is less equal than Britain. So too is Russia, Turkey, China and the United States.

20

Wealth inequality: UK trends

As with income inequality, the trends and level of wealth inequality often do not conform to the narrative expressed in the media.

The ONS’s ‘Wealth and Assets Survey’ provides summary statistics of the level of household wealth in the UK for 2006-08, 2008-10, 2010-12 and 2012-14. It calculates household wealth by summing up net property wealth, physical wealth, net financial wealth, and private pension wealth from household survey data, allowing us to calculate a host of summary statistics of inequality, including the Gini coefficient and the wealth share of various groups in the distribution.2

Table 2: Measures of wealth inequality

Source: ONS (2015)

As can be seen, the Gini coefficient at 63 is significantly higher than for income – a fact that tends to be true across all major countries (Credit Suisse 2014). This makes sense, not least because of the life-cycle effect (people just entering the labour force tend to have no assets but do have incomes, for example).

2 Net property wealth in this instance means any property wealth less the value of any mortgages held against the properties. Net financial wealth includes money saved in formal financial assets, money saved under the bed (informally) and children’s assets minus any financial liabilities held.

21

What is clear from this ONS data though is that wealth inequality has been fairly stable over the four waves considered, on almost all measures. The Gini coefficient has been almost completely flat; the share of total wealth attributable to the top ten per cent has risen by almost 1 percentage point but the share of the top 1 per cent has fallen slightly. The share of wealth attributable to the bottom 50 per cent was completely stable for the first three waves but then has fallen somewhat in the last. It’s only when you get to the top 0.1 per cent that you find the share of total wealth for those at the top has increased significantly in recent years, though when you are looking at this group you are talking about incredibly small numbers of observations in these surveys which make the results highly unreliable and perhaps not very meaningful.

The 2012/14 data do show that, very recently, wealth inequality has risen for the first time in a decade (Giles 2015). But this has overwhelmingly been driven by net property wealth, following rising house prices in London and the south east of England in particular. There is a host of academic evidence that the primary cause of this long-term structural phenomenon is land-use planning regulation, which somewhat counters the case often heard that the main policy response needed to lower wealth inequality is higher taxes (Hilber and Vermeulen 2016).

One additional problem with these crude wealth inequality statistic is that they ignore some state-provided entitlements such as state pensions or social security, which in essence are wealth entitlements, albeit in the form that other taxpayers will pay for them. Their inclusion would considerably flatten the wealth distribution (because they are an essentially fixed endowment) as would state health provision flatten the income distribution. Of course, pensions eventually come out in the income stats when the pension is received.

What about wealth inequality in the longer term? Thomas Piketty’s best-selling book ‘Capital in the Twenty First Century’ claims both that wealth inequality is much higher than the ONS figures suggest and that there has been a rise in wealth inequality in the UK since the 1980s (Piketty 2014a).

22

Figure 4 below shows Piketty’s historical series. The raw numbers from which Piketty constructed his series are shown in blue (Giles 2014a). His data show that wealth inequality (as measured by the share of total wealth attributable to the top ten per cent or top one per cent) fell significantly from the First World War through to the 1970s, and is still much lower today than the vast majority of the period since 1810. It is what has happened since 1970 which is the subject of intense debate (see Piketty 2014b, Giles 2014a, Giles 2014b).

Figure 4: Piketty and Piketty sources on wealth share of top 10% and top 1%

Source: data from Chris Giles’ spreadsheets; top lines show top 10%, bottom top 1%.

Piketty’s analysis suggests that the top ten per cent of the wealth distribution had as much as 70.5 per cent of UK wealth and the top one per cent as much as 28 per cent in 2010. These are much higher than the equivalent ONS figures outlined above. Piketty argues that survey data – as used by the ONS – tends to significantly understate the wealth of those at the very top. Thus, he prefers to use HMRC tax return figures for inheritance for the 2010 observation, despite this data source being criticised as not suitable for this calculation. Use of this observation allows him to claim that overall inequality has increased since the 1980s and much higher the ONS imply.

23

There appears to be some major differences between Piketty’s series and the data from Piketty’s sources for the interim period (Giles 2014a, 2014b). The source data series seem to show a significant fall in wealth inequality through the 1970s, whereas Piketty’s constructed line shows a very modest fall. For the 1980s, the data from the sources suggest a much lower level of wealth inequality.

The main problem seems to be that there are discontinuities in the data (i.e. when Piketty has to shift from one data source to another which cross over, there are significant differences in the level of inequality for the same year – see Reed 2014). Some have suggested that upward adjustments therefore need to be made to more recent data to make it historically comparable to the older series. These adjustments are calculated to be highly significant - by 2010, as large as 23 percentage points for the top ten per cent, and ten percentage points for the top one per cent (Reed 2014).

Figure 5: Piketty and Howard Reed series adjusting for discontinuities

Source: Reed (2014)

24

It’s unclear why it is more sensible to revise data upwards to make it comparable with older series rather than revising down the older series. And Piketty does not explain his series by appealing to discontinuities. Nevertheless, making these adjustments does lead to results similar to Piketty’s (see Figure 5), though doing this shows wealth inequality is still slightly lower than Piketty suggests and was flat between 2000 and 2010 (the modest increase almost all comes in the 1990s).

Whilst there is a debate about the level of wealth inequality stemming from Piketty’s work, it is clear that wealth inequality fell substantially for most of the 20th century, and has risen either very modestly or remained essentially stable since the 1980s. In other words, levels of wealth inequality are neither unprecedented nor exploding. Giles concludes that the level of wealth inequality is likely to be a bit higher than suggested from the Wealth and Assets Survey but nowhere near as high as Piketty suggests.

25

Wealth inequality: international comparisons

Due to the data difficulties outlined above, it is very difficult to make international comparisons of wealth inequality across countries. Nevertheless, Credit Suisse’s Global Wealth Databook attempts to achieve just that. Their net wealth measure examines ‘the marketable value of financial assets plus non-financial assets (principally housing and land) less debts’.

This throws up some perverse and counterintuitive results in some countries (discussed in the global inequality section below). Their data has also been criticised for being spliced from a number of different sources (Giles 2016).

Nevertheless, this dataset is frequently cited by Oxfam and is one of the few sources which compares wealth inequality across a range of developed countries.

The UK’s Gini coefficient for net wealth in this source is calculated to be 68.2 - the sixth lowest of 25 countries for which there is data (see Figure 6). The net wealth shares for the top ten per cent and top one per cent are 54.1 per cent and 23.3 per cent – meaning the UK is sixth and eighth lowest out of 22 countries on these measures (see Figures 7 and 8). The UK also has the sixth highest figure for net wealth attributable to the bottom 50 per cent of the distribution of 21 countries for which there is data (8.4 per cent of total wealth).

These cross-country statistics showing that the UK’s net wealth inequality is not as high as in other countries are barely mentioned when campaigning groups such as Oxfam extract their findings from the Credit Suisse reports.

26

Instead, Oxfam highlights the level of overall wealth inequality at a global level, before looking at trends in net wealth inequality in the UK – which, unlike the other wealth figures above, do suggest rising inequality (but still to a much lower level than countries such as Sweden and Denmark (Credit Suisse 2014)).

Figure 6: Gini coefficient for wealth inequality for developed countries with data

27

Figure 7: Net wealth share of the top ten per cent of the wealth distribution

Figure 8: Net wealth share of the top one per cent of the wealth distribution

Source: Credit Suisse (2014)

28

Global income and wealth inequality

When discussing income inequality, it is common to focus on within-country changes or levels, as discussed in the sections above. Those who purport to highlight the plight of the world’s poorest on a global level become oddly nationalistic when discussing the distribution of income.

Yet there is plenty of evidence suggesting that income inequality has been falling at a global level, or at least not rising, in the past two decades. Lakner and Milanovic (2014) show that the income distribution has become more compressed in the last twenty years, largely due to rapid economic growth in many Asian economies (such as China, India and Indonesia) which has lifted hundreds of millions out of poverty and reduced the income gap between these countries and richer countries.

In fact, they tentatively conclude that since 1988 the forces of globalisation have led to significant rises in income for many in these countries which may well be associated with a slower growth of incomes for those in the lower deciles of some richer countries. The top one per cent globally have also seen very rapid income growth – leading to a more homogenised ‘global rich’, less linked to any particular country. Though it is difficult to track accurately the incomes of those at the very top, they conclude that the past 30 years have probably equalised incomes to a greater extent than anything since the Industrial Revolution.

There are all sorts of data issues which make producing these global income distribution figures difficult. For example, incomes of the very richest tend to be understated by ordinary household surveys and there is a need to adjust for purchasing power parity to account for different price levels across countries. Some studies, such as Anand and Segal

29

(2014) show that, on some measures, there was an increase in inequality expressed through the income share of the global one per cent prior to the financial crisis. But broadly, most of the literature in this area suggests that income inequality (while being much larger than at most within-nation levels) has fallen globally over the past generation.

Perhaps due to this, it is common for egalitarian campaign groups to instead highlight apparently ‘shocking’ statistics in relation to the level of global wealth inequality. Oxfam, in particular, has been keen to use the Credit Suisse data on global wealth to claim that ‘the world’s 85 richest people own the same wealth as the bottom 3.5 billion combined’.

These figures produce a hugely distorted picture of the gap between ‘rich’ and ‘poor’. They refer to net wealth data – which measures ownership of assets minus debts; and a range of adjustments to the data are made to estimate the wealth held by the very richest.

There are two main problems with using global distributions of net wealth to conclude anything about global inequality. The first is that it should not really surprise us that those right the way up to the middle of the global income distribution have barely any net assets (an adult needs only $3,210 in assets, once debts have been subtracted, to be in the top half of the distribution). The median age of the world population is just 29.7 years old; and the median age of the adult population is between 35 and 39. From a lifecycle perspective, one would not expect people to have much in the way of net assets up to around this age, even in rich countries.

Second, and perhaps more importantly, examining net wealth figures implies that China has no people in the bottom ten per cent, with most Chinese in the top 50 per cent, whereas North America supposedly has around eight per cent of the world’s poorest people. This is because significant numbers of people in the US are loaded up with debts of various kind, making their net wealth negative.

Whilst there is nothing inherently wrong with interpreting figures in this way, it is doubtful that many would consider, say, a Harvard graduate with no assets but high debts to be one of the world’s poorest people. High levels of debt can also be a sign of prosperity in that only people with good repayment prospects can be expected to be given significant loans. Summing a lot of negative or zero figures together of this kind to find out how much the top one per cent or top ten per cent ‘own’ in net wealth

30

terms relative to a bottom chunk of the global population is always going to lean towards these striking results.

Oxfam has attempted to temper this criticism by presenting figures excluding debts which they say produces similar results. Yet the demographic critique remains unanswered. It is therefore very difficult to conclude anything meaningful about global wealth inequality from the data sources available.

31

How should we think about inequality statistics?

Thus far this paper has assessed inequality in a positive sense – examining objective facts. But the political debate about inequality is often a normative one. Many people start from the value judgement that more inequality is a bad thing and less inequality is a good thing per se. This moral outlook in favour of less inequality is then used to justify government action to narrow an economic distribution.

Proponents of this view often describe a distribution of income or wealth after-the-event as if it is pre-determined or distributed. Indeed, even the term ‘distribution’ implies that income or wealth is allocated to different groups from some central source. Yet inequality measures are in reality summary statistics which reflect the results of millions of individual interactions, exchanges, endowments and policies. Though they give us aggregated information, they tell us nothing about how that distribution has arisen, or any trends we might have seen. If one does not subscribe to the egalitarian worldview, but instead starts from the view that the morality of a distribution depends on whether its causes are just, then the summary statistics analysed in the previous sections are not particularly helpful as a guide to policy.

There is also a tendency when discussing inequality of income or wealth to describe them rhetorically as if there is a fixed pie of either – and that one person having income or wealth must come at the expense of somebody else. Many papers, even from academic experts, talk of people (usually the rich) ‘extracting’ wealth or ‘claiming’ it.

In reality, economic growth allows everybody to get better off over time. In a society with no growth, one person’s gain would inevitably be another

32

person’s loss; the rich could only get richer by making the poor poorer. Fortunately, the economy is not a zero-sum game. People can and often do get rich as a result of entrepreneurial activity that makes everyone better off. Of course, some people get rich off the back of crony capitalist activity as well – which really can be zero sum - but there is no generalised rule that wealth or income obtained is exploitative or unjust, as many imply.3

Moreover, income and wealth are not distributed by some higher power. They are largely a return for some economic activity – whether labour or investment. If money was handed out at random to individuals in large and small amounts regardless of their effort or merit, it would indeed be unfair. As Hayek (1998: 228) wrote in Law, Legislation and Liberty, ‘the manner in which the benefits and burdens are apportioned by the market mechanism would in many instance have to be regarded as very unjust if it were the result of deliberate allocation to particular people. But this is not the case.’ In a free market, the ‘distribution’ of income depends on the value of an individual’s labour or other activity to others.

This fact often gets lost in debates about this issue, where market ‘worth’ is blurred rhetorically with our worth as humans. Many egalitarians like to describe riches as ‘unfair’. ‘Surely it is morally wrong that a footballer earns so much more than a teacher or nurse?’ they say. It is tempting in response for those sceptical about state attempts to impose equality to swing the argument the other way and to claim that all returns to labour or investment are earned or just. In fact, market exchange is determined by interactions of supply and demand – systems of exchanges which require no moral judgements. All sides should acknowledge that even if we just consider this, some individuals do get into certain positions as a result of good fortune or unequal opportunities. The real argument against an egalitarian state agenda is that attempts to correct for inequalities entail entrusting a central source in policing what is moral or just, often with huge unforeseen consequences and damaging effects on incentives.

We should bear all this in mind when talking about what the government should to do ‘curb inequality’. Affecting a distribution of income or wealth inevitably means interfering with some of the trades or interactions which

3 In academia, university income really is a ‘fixed pie’ to be fought over through grant applications. The same is true of the public sector, where fixed annual budgets are distributed by a central power after inter-departmental negotiations, thereby creating winners and losers. One can only speculate about whether those who work in these sectors have their views of the wider economy coloured by such experiences.

33

lead to a given set of outcomes. Government cannot control inequality directly. It can only seek to affect the eventual distribution by changing human actions.

A big concern is that interfering extensively in these interactions to lower inequality might in some instances also have adverse consequences for economic growth. The logic of saying you are in favour of reducing inequality as your primary aim suggests you would be willing to accept a more impoverished society as a whole, provided that lower level of riches are more evenly distributed.

Of course, at extremes it is clear that total inequality (one person having everything) and total equality (everybody having the same) are undesirable. If everything was owned by one person then everybody else would obviously be in desperate poverty. On the other hand, if everybody was guaranteed the same share of national income, nobody would have an incentive to work and there would be no income to distribute. Moreover, both societies would require violence, oppression and the abolition of markets.

If both equality and inequality are deplorable in extremis, it would be tempting to think there existed a sweet spot on the spectrum between them. But if egalitarians believe that there is a socially optimal level of inequality, they have never specified what it is. We doubt that such a level exists. It is, in our view, impossible to say whether inequality is too high (or too low) in a given country without knowing the reasons behind it.

Inequality is simply an economic indicator; one of many. What it indicates can be good, bad or indifferent, but it cannot be good or bad in itself. Extreme levels of inequality, as seen in countries such as South Africa, can be indicative of historical injustices. They can result from prejudice and oppression, past and present. They can result from government capture by special interest groups or by cronyism and corruption. They can result from a lack of democracy.

All of these negative causes are undesirable in and of themselves. Poor education, family breakdown, racial discrimination, long-term unemployment and social immobility might all fuel inequality but they are all evils in themselves which should be dealt with regardless of what the Gini coefficient says. They would be undesirable even if they did not breed inequality (and often they do not).

34

On the other hand, there are some causes of inequality which are benign, such as lotteries, or positively beneficial, such as immigration, technology and free trade. In South Africa, for example, income inequality actually rose post-apartheid, as talented blacks had opportunities open to them for the first time and the economy as a whole became more dynamic as the political institutions of the country changed (Acemoglu and Robinson 2015).

Starting from a position where more inequality is always a bad thing might therefore lead someone to pursue policies which might actively be harmful to society overall and particularly to those on low incomes. As an example, we could lower inequality (at least temporarily) by deporting or exterminating rich people. But would this benefit those who remained? It is difficult to see how it would.

Britain has seen a significant fall in income inequality in recent years thanks to a sustained fall in wages caused by a serious financial meltdown. China, by contrast, has seen rising inequality as a result of increased productivity and prosperity. In Britain, greater equality has been a symptom of a problem. In China, greater inequality has been a symptom of success. It would take an extreme relativist to argue that Britain has become better and China has become worse.

Of course, we do not pretend that inequality is always reflective of ‘good’ trends. Our point here is that it cannot be generalised that less inequality is necessarily better, as often implied. If inequality is merely an indicator then efforts to improve society by reducing the gap between rich and poor are akin to trying to make a dog happy by wagging its tail. Summary statistics simply do not give us enough information.

35

Can inequality itself be the cause of problems?

Inequality statistics alone tell us little about the desirability or otherwise of a distribution of income. Without knowing the causes of that distribution, the logic of taking active steps to narrow a distribution might mean a willingness to accept other undesirable outcomes. As an example, Bill Gates got wealthy as a result of providing goods and services which hugely enriched society. His success meant he became extremely rich, and increased overall inequality, but yet his personal fortunes were dwarfed by the value and utility he provided to his customers and the wider world. His activities made computer operating systems available to many more people at affordable prices, reducing consumption inequality too. Preventing or disincentivising him from undertaking those activities might have reduced inequality, but the damaging consequences for society would have been great indeed.

However, over the past decade it has been suggested that inequality may be more than an indicator. Some claim instead that inequality can be the direct cause of future economic and social problems. Broadly speaking, there are three hypotheses.

1. Inequality causes bad social outcomes

The first hypothesis says that the psychological impact of inequality causes negative social outcomes, including ill health. This theory was developed by the sociologist Richard Wilkinson and was popularised in The Spirit Level, a book he wrote with the epidemiologist Kate Pickett in 2009. One of us has written at length about The Spirit Level hypothesis in previous IEA publications so we will not give a full critique here (Snowdon 2012, 2015; see also Snowdon 2010).

36

In short, the book uses social science to ‘prove’ that the prevalence of health and social problems in a society is highly dependent on the level of income inequality. A series of bivariate scatterplots are used to show that ‘less equal’ countries have higher rates of, for example, murder and obesity than do ‘more equal’ countries. Similar correlations are shown for US states. To explain these statistical relationships, the authors suggest that income inequality erodes social capital and creates psychological stress, thereby leading to violence, ill health and a breakdown of trust.

The main problem with The Spirit Level is the statistical evidence itself. Most of the associations between inequality and social problems are weak and many of them are driven by one or two outliers at the extreme ends (the USA and Japan, in particular). Several countries are excluded from the international analyses which, if included, make most of the correlations disappear entirely. Statistical relationships are even weaker in the analyses of US states but insofar as correlations exist they are largely driven by the fact that the worst performing states which happen to be the least equal are also the poorest.

There is a large body of evidence suggesting that Wilkinson and Pickett’s theory is incorrect, though they acknowledge none of it in the book. The claimed (inverse) relationship between inequality and life expectancy has been the subject of a great deal of research over three decades but, as the Oxford Handbook of Economic Inequality notes, ‘most studies of health and inequality find no statistically significant relationship either across countries or over time’ (Leigh et al. 2009). As a ‘theory of everything’, The Spirit Level’s simple hypothesis would have been a breakthrough in human understanding had it been true, but there is ample evidence that it is not.

2. Inequality corrupts the democratic process

The second hypothesis says that income inequality leads to the corruption of the democratic process. Thomas Piketty argues that high levels of inequality can ‘lead to a capture of the political process by a tiny high-income and high-wealth elite. This directly threatens our democratic institutions and values.’

But the idea that the super-rich are a homogenous group of right-wing ideologues who buy elections on a whim is not consistent with the facts. In his review of Piketty’s book, Jonah Goldberg surveyed the American

37

political landscape and noted: ‘At this very moment, George Soros, Tom Steyer, and other liberal billionaires are in a hammer-and-tongs political battle with Sheldon Adelson, Charles and David Koch, and other conservative or libertarian billionaires.’ American democracy, says Goldberg, is not being run by robber barons and aristocrats. ‘It’s being run, instead, by the son of a teenage single mother from Hawaii [Barack Obama], the son of a barkeep from Ohio who became speaker of the House [John Boehner], and a miner’s son from Nevada who grew up in a shack with no running water before becoming majority leader of the Senate [Harry Reid]—none of them born into wealth, to put it mildly’ (Goldberg 2014).

It is well known that David Cameron went to Eton College, but he is the first old Etonian to be prime minister since 1964 and is one of only two prime ministers in the last fifty years to have gone to public school (the other is Tony Blair). If Conservative members had voted for David Cameron’s rival David Davis to become leader of their party, Britain may now have a prime minister who was brought up on a council estate by a single mother.

In 2015, the proportion of Conservative MPs who had been privately educated fell below 50 per cent for the first time. Overall, 33 per cent of MPs elected in 2015 had been privately educated, down from 49 per cent in 1979. The proportion of Oxbridge-educated politicians has fallen by around the same amount (49 per cent to 34 per cent) (Sutton Trust 2015).4

This is not to say that money has no effect on politics, but the issue that concerns critics such as Piketty is not wealth per se but wealth inequality, which is to say the accumulation of wealth by those in the highest income groups. Since very wealthy people have existed throughout the history of democracy, it is not clear why it matters whether Donald Trump and Michael Bloomberg are worth $5 billion or $25 billon. The mechanism by which modern inequality, rather than traditional wealth, corrupts democracy has never been adequately described.

Power is ultimately in the hands of the electorate. ‘We are the 99 per cent’ is the slogan of the neo-egalitarians. If so, this confers an obvious numerical advantage in elections. It is patronising to assume voters elect politicians because they have fallen victim to expensive political advertising campaigns. In any case, the link between campaign spending and electoral success is tenuous, at best (Levitt 1994).

4 Only Conservative, Labour and Liberal Democrat MPs have been included in this analysis. Proportions are significantly lower among SNP politicians.

38

3. Inequality slows economic growth

The third hypothesis says that inequality slows economic growth. There have been numerous different mechanisms put forward to examine why this might be, but the empirical literature is extraordinarily mixed and often contradictory. Some studies find a negative relationship between income inequality and growth (Ostry et al. 2014; Berg and Ostry 2011), some find a positive relationship (Forbes 2000), some find no correlation or relationship at all (Krugman 2015), and others find a negative correlation between changes in the level of inequality and growth in any direction (Banerjee and Duflo 2003).

Perhaps because of this muddy empirical evidence, many have sought to identify potential mechanisms by which inequality and growth could be related.

The first mechanism is one cited by Raghuram Rajan (2012) but variants of the story are also told by many progressive economists too. This says that rising income inequality in the USA and elsewhere had the effect of leading to expanded credit being pushed further down the income scale in the build up to the financial crisis – and this became a key cause of the crash. For some economists, this was the result of the rich’s higher marginal propensity to save – leading to a larger pool of available funds and pushing down borrowing costs for people lower down the income scale. For others, such as Rajan, the mechanism is political. Political concern about those left behind led a desire to make credit available to those lower down the income scale – exacerbating the sub-prime crisis. Demand for credit in both instances came from those on lower incomes desire to ‘keep up with the Joneses’ in the face of their own stagnating incomes.

The first thing to note about this thesis is that it does not appear to be able to explain similar experiences on both sides of the Atlantic. In the UK, as we have seen, inequality was largely stagnant in the period from the 1990s to the financial crisis, but the incomes of the poor were growing steadily. In the US, inequality increased somewhat whilst the incomes of poorer groups remained stagnant. In both countries, saving fell significantly and household debt increased. This suggests that other factors, such as the global savings glut, monetary policy and rising house prices in both countries may have been more significant factors in leading to the conditions which enabled the financial crisis than either inequality or the public policy response to it.

39

The second thing to note is that empirical studies attempting to test this ‘keeping up with the Joneses’ demand aspect of the hypothesis have found no evidence that it is true. Coibin et al. (2014), for example, tested the thesis by examining the relationship between regional inequality and household debt in US regions, and actually found that low-income households in high-inequality regions borrowed less than those in low-inequality regions. This suggests rising inequality would actually lead to a lower amount of borrowing.

Sadly, lots of people seem to use motivated reasoning to attribute rising inequality to financial crises, yet most academic analyses find no systematic relationship between inequality and credit booms (Bordo and Meissner 2012).

A second mechanism through which rising inequality has been said to deter growth has been proposed by the OECD. They believe that a wider dispersion of income might deter growth by depressing skills development among individuals with poorer educational backgrounds, both in terms of the quantity and quality of education obtained. This theory was used as an explanation for results from a cross-country analysis they undertook, which was widely reported as suggesting that higher inequality slowed economic growth (Cingano 2014). Inevitably, the research concluded that redistributive policies were necessary to encourage growth.

Yet the OECD’s conclusions were not as strong as the media coverage suggested. It explicitly found ‘no evidence … that those with high incomes pulling away from the rest of the population harms growth’. The inequality they identified as problematic for growth was larger dispersions in the bottom half of the income distribution. This seems more a story about poverty than inequality.

And their measure of redistribution, they readily admit, is ‘crude’. Indeed, if their mechanism is correct, we would expect education and welfare reforms to be far more potent solutions to the human capital accumulation problem than redistribution. To the extent that their redistribution recommendations could work, given their results, more redistribution should occur within the bottom half of the income distribution than from the richest to the poorest. It is unclear that this is what those concerned with inequality have in mind, though other papers have suggested that tax policies which lower inequality in the bottom half of the income distribution can improve growth (Biswas et al 2015).

40

The literature on this subject is therefore mixed, as many progressive economists acknowledge (Krugman 2015). Since economists have traditionally seen a trade-off between equalising incomes and fostering growth, some left-leaning academics settle for arguing that the evidence shows redistributive measures to reduce inequality might not reduce growth too much. This may be true, up to a point.

The IMF has, in the past, published research which suggests an inverse-U relationship between the degree of redistribution and economic growth – i.e. that low levels of redistribution can aid economic growth but further redistribution harms it (Ostry at al 2014). According to its analysis, most OECD countries, with the exception of the USA, are already beyond the level of redistribution at which growth is maximised.

In summary, there is little evidence of a relationship between inequality and economic growth, and where a correlation has been identified, no mechanism from high inequality to low growth has been tested as robust. The idea that inequality slows the growth in living standards seems a good example of motivated reasoning from egalitarians.

41

Conclusion

The tale we are told about inequality in Britain is more fiction than fact. Again and again it is claimed that the UK is being crippled by spiralling inequalities of income and wealth, and that the rich are getting richer while the poor get poorer. It is ironic that this narrative has come to the boil in Britain since 2008 when income inequality has been in decline and the rich have seen their incomes fall by more than any other group. Even before the financial crisis, inequality had not risen for many years and Britain, though less equal than many EU countries, was neither more nor less equal than the rest of the world.

Given the amount of exaggeration and misinformation surrounding the issue of inequality in Britain, it is not surprising that people appear to be increasingly concerned about it. The very word ‘equality’ brings to mind the inarguably positive concepts of equal rights and equality before the law. Who wishes to be thought in favour of inequality? (Not us. We merely think it unimportant.) And yet the British have always been fonder of equal rights than equal outcomes (or, indeed, equal incomes). When Ed Miliband put the issue of income equality at the heart of his 2015 election campaign, it did not end well for him. People may support greater equality in the abstract without supporting the policies that would make it happen.

Perhaps it is not inequality they are concerned about but something else. When the Work Foundation conducted surveys in four British cities in 2013, many respondents expressed concern about inequality, but when pressed further it became clear that ‘there was a tendency to see the concepts of inequality and poverty as the same thing… expressions of concern about inequality were often actually concerns about poverty’ (Lee et al. 2013: 35).

42

The distinction between poverty and inequality is critical. The gap between high and low earners tells us nothing about the material living standards of the poor, and a preoccupation with the super-rich will do nothing to help people on low incomes. No doubt there are some policies put forward under the banner of reducing inequality that would have positive effects on the living standards of the poor, but many would be of no benefit to anybody and some would directly harm the poor by, for example, creating unemployment and stifling growth.

The inequality agenda harnesses the almost universal concern for people who are worse off than ourselves and twists it into anger at those who are better off. It transforms compassion into resentment, as if the economy were a zero sum game; as if wealth was the cause of poverty. This is a great mistake. The economy is not a fixed pie and there is no direct connection between poverty and inequality. Indeed, the relationship between the two is often inverse, with rates of inequality falling during periods of economic hardship such as the post-2008 slump, the recessions of the early 1990s and mid-1970s, and during both world wars.

Spirit Level egalitarians believe that making the rich poorer would, in itself, bring benefits to the poor even if the poor became no richer. This, admittedly, is the extreme relativist view. The more mainstream left-wing view is that you reduce inequality by taking from the rich and giving to the poor. If so, reducing inequality is a really just a byproduct of achieving the core aim of alleviating poverty. But if reducing inequality is just a euphemism for reducing poverty, it opens the question of whether greater wealth distribution is, in fact, the best way to accomplish it. A debate can be had about whether benefits are too generous or too stingy. We can discuss reform of the welfare state and argue about how much tax revenue the government can realistically expect to receive. We can make plans to improve education and social mobility. But such debates should take place on their own terms and not be influenced by what effect a given policy might have on one economic indicator.

The pressing economic questions of our time remain the same as they have always been: how to grow the economy, increase pay, create jobs, control inflation and live within our means. Traditional socialism had no good answers to these questions and it has no better answers under its new guise of egalitarianism. By obsessing about inequality, the left is rewording the question in a way that makes socialism an acceptable answer. If reducing the gap between rich and poor is the overarching goal

43

of public policy, if it is more important than economic growth, and if the purported psychological benefits of greater equality are more important than material living standards, then a dirigiste economy of high taxes and low growth becomes more appealing.

We hold that it is better to be materially well off than relatively well off. As mentioned above, some egalitarians are strangely parochial in their concern about inequality. Taken as a whole, the world has become more equal in recent decades, largely as a result of China and India becoming richer. In China, unprecedented economic growth has gone hand-in-hand with a growing gap between rich and poor. This creates a dilemma for the egalitarians. Should they be pleased that the world is a more equal place or should they be dismayed that China has become less equal? Perhaps they are able to entertain both emotions simultaneously, pitying the Chinese while feeling good about the rest of the world.5 For our part, it seems clear that the new wealth and opportunity afforded to the Chinese through economic growth is the real prize to be celebrated. Whether their improved standard of living has increased or reduced inequality - or done both under different measures - is neither here nor there.

At worst, inequality can be a symptom of a problem, but it is not a problem per se. You do not solve social problems by tackling symptoms. The causes of inequality can be positive or negative, just as efforts to reduce inequality can produce positive or negative outcomes. The problem with egalitarianism is that it pursues a single economic variable for its own sake regardless of its effect on other outcomes (or, in the case of Spirit Level egalitarianism, on the assumption that it will somehow improve other outcomes). Much better to focus on the outcomes that matter and address them directly.

If it was a straight choice between higher wages and greater inequality, we would gladly take the higher wages. We would accept greater inequality if that was the price of getting more people into work or having more mobility between the classes. To be clear, we are not claiming that these are always necessarily the trade-offs that have to be made but if they were, we know which we would choose. It is not always clear what the egalitarians would choose.

5 In practice, they do neither. Characteristically ignoring the statistics, they simply assert that the world is becoming a less equal place.

44

References

Acemoglu, D. and Robinson, J. (2015) The Rise and Decline of General Laws of Capitalism. Journal of Economic Perspectives Volume 29, Number 1. Winter 2015: 3–28.

Banjeree, A. and Duflo, E. (2003) Inequality and Growth: What Can the Data Say? Journal of Economic Growth. September 2003. Volume 8, Issue 3: 267-299.

Berg, A., and Ostry, J. (2011) Inequality and Unsustainable Growth: Two Sides of the Same Coin? IMF Staff Discussion Note. August 2011. International Monetary Fund, Washington.

Biswas, S., Chakraborty, I. and Rong, H. (2015) Income Inequality, Tax Policy, and Economic Growth. Becker Friedman Institute for Research in Economics Working Paper No. 2595524. July.

Bordo, M. and Meissner, C. (2012) Does Inequality Lead to a Financial Crisis? NBER Working Paper No. 17896. March.

Cingano, F. (2014) Trends in Income Inequality and its Impact on Economic Growth. OECD Social, Employment and Migration Working Papers. No. 163. OECD Publishing.

Coibion, O., Gorodnichenko, Y., Kudlyak, M. and Mondragon, J. (2014) Does Greater Inequality Lead to More Household Borrowing? New Evidence from Household Data. NBER Working Paper 19850.

Credit Suisse (2014) Global Wealth Databook 2014. Credit Suisse Research Institute. October.

45

Cribb, J. (2013), ‘Income inequality in the UK’, London: Institute of Fiscal Studies.

Forbes, K. (2000) A Reassessment of the Relationship between Inequality and Growth. American Economic Review 90(4): 869-887.

Giles, C. (2014a) Data problems with Capital in the 21st Century. Ft.com Money Supply blog. 23 May.

Giles, C. (2014b) Capital in the 21st Century – a response. Ft.com Money Supply blog. 30 May.

Giles, C. (2015) UK wealth inequality widens for first time in a decade. Financial Times. 18 Dec.

Giles, C. (2016) Three reasons to question Oxfam’s inequality figures. FT Data blog. 18 Jan.

Hassan, G. (2012) The fourth most unequal country in the world. Scotsman 17 March.

Hilber, C. and Vermeulen, W. (2016) Regulation is to blame for England’s surging house prices. VOX CEPR’s Policy Portal. 10 Apr.

Krugman, P. (2015) Musings on Inequality and Growth. The Conscience of a Liberal blog. New York Times.

Lakner, C. and Milanovic, B. (2013). Global income distribution: from the fall of the Berlin Wall to the Great Recession. World Bank Working Paper No. 6719. December.

Lee, N., P. Simmons and K. Jones (2013) Wage inequality and employment polarisation in British cities. Work Foundation

Leigh, A., Jencks, C. and Smeeding, T. M. (2009) Health and economic inequality. In The Oxford Handbook of Economic Inequality (ed. W. Salverda, B. Nolvan and T. M. Smeeding) Oxford University Press: 384–405.

Levitt, S. (1994) Using repeat challengers to estimate the effect of campaign spending on election outcomes in the U.S. House. Journal of Political Economy 102(4): 777-98

46

Office for National Statistics (2012) Real wages up 62% on average over the past 25 years. 7 November

Office for National Statistics (2014) Wealth in Great Britain Wave 3, 2010-2012. London: Office for National Statistics

Office for National Statistics (2015a) The effects of taxes and benefits on household income, financial year ending 2014. 29 June

Office for National Statistics (2015b) Wealth in Great Britain Wave 4, 2012-2014. London: Office for National Statistics

Office for National Statistics (2016) The effects of taxes and benefits on income inequality: 1977 to financial year ending 2015. 8 April

Ostry, J., Berg, A. and Tsangarides, C. (2014) Redistribution, Inequality, and Growth. IMF Staff Discussion Note. February 2014. International Monetary Fund. Washington.

Oxfam Twitter feed. 5 June 2014 https://twitter.com/oxfamgb/status/474448271596060672

Piketty, T. and E. Saez (2012), ‘Top incomes and the Great Recession: Recent evolutions and policy implications’, 13th Jacques Polak Annual Research Conference

Piketty, T. (2014a) Capital in the Twenty-First Century.

Piketty, T. (2014b) Technical appendix of the book. Capital in the twenty first century. Appendix to chapter 10. Inequality of Capital Ownership. Addendum: Response to FT.

Rajan, R. (2010) Fault Lines: How Hidden Fault Lines Still Threaten the World Economy. Princeton, NJ: Princeton University Press.

Reed, H. (2014) Piketty, Chris Giles and wealth inequality: it’s all about the discontinuities. Guardian UK News Datablog. 29 May.

Russia Today (2015) Inequality Street: UK most unequal country in the EU, worse than US. 19 May

47

Snowdon, C. (2010) The Spirit Level Delusion. Ripon: Little Dice

Snowdon, C. (2012) ‘Are more equal countries happier?’ in …and the Pursuit of Happiness (ed. P. Booth). London: Institute of Economic Affairs

Snowdon, C. (2015) Selfishness, Greed and Capitalism. London: Institute of Economic Affairs

Sudhir, A. and Segal, P. (2008). What Do We Know about Global Income Inequality? Journal of Economic Literature. 46(1): 57–94.

Sudhir, A. and Segal, P. (2008). The Global Distribution of Income. International Development Institute Working Paper 2014. 01.

World Bank. GINI index (World Bank estimate: http://data.worldbank.org/indicator/SI.POV.GINI (accessed 25 January 2016)

48

Appendix: Piketty’s thesis and criticism

Thomas Piketty’s ‘Capital in the Twenty First Century’ sought to analyse long-term trends in inequality and argued that in future we might see dynamics which result in high and rising inequality within countries. He believes that capitalism itself has a tendency to produce large and unsustainable inequalities, necessitating activist policy responses, such as a global net wealth tax and higher marginal income tax rates on top earners.

His evidence and thesis for this can be summarised succinctly. His data collection concludes that total income inequality (labour income and income streams as returns from assets) was very high in the US and Europe prior to the First World War. The two world wars and their aftermath then led to a huge erosion in the importance of capital as wealth was destroyed both by conflict and policy responses, such that inequality fell substantially. Since then, as countries have liberalised their economies, income inequality and wealth inequality have increased – particularly in the United States. This broad sweep of history, Piketty suggests, shows how capitalism uninterrupted by conflict and activist state policies tends to lead to high levels of inequality.

Piketty expects levels of inequality to rise in the future for two reasons. The first is what economists might describe as a ‘capital share’ effect. Piketty believes economic growth will slow substantially primarily due to changing demographics and lower measured productivity growth. Over time, if the savings ratio in the economy does not change much and growth is sluggish, theory would suggest an increase in the ratio of accumulated wealth to national income. Piketty assumes that this higher capital to income ratio would not lead to a substantial fall on the returns to capital (because he believes that the elasticity of substitution between capital

49

and labour is greater than 1.0), such that the capital share of national income would rise. If the share of income going to capital increases, and capital is very unevenly distributed (which it is), then we would expect that national income itself would become more unevenly distributed given these conditions. Therefore the level of inequality will increase.

The second ‘capital concentration’ effect relates to the relationship between the rate of return on capital and the growth rate of the economy, and the effect this has on increasing concentrations of wealth. The essential insight is that people who obtain a large fraction of their income from returns to capital at any given time are more likely to see their position entrenched if the average returns on capital (r) are higher than the growth rate of the economy (g). In simple terms, if r < g, then even if a family used all their income for accumulation of capital, the family’s returns from capital over time would grow more slowly than the economy so their position would erode. If r > g, then dynastic wealth can grow significantly, provided savings and inheritance behaviour cooperate. Piketty’s argument is that in most times r > g, but that in future the gap between r and g is likely to increase even further due to slower economic growth. This is an economic force that can lead, ceteris paribus (i.e. provided assumptions are made about savings and inheritance transmission) to ever-rising inequality.

Piketty’s work certainly attracted lots of attention upon release. Initially it received many favourable reviews, with criticisms mainly focused on his reticence to explain why increased inequality was in itself a problem (beyond some generalised claims about democratic implications, explored later in this paper) and his bold ‘solutions’, such as a global net wealth tax, which were seen as utopian and unworkable.

Perhaps because of his academic prominence, Piketty’s work then attracted much more high-calibre scrutiny than many other more populist works on inequality. Even leaving aside the debate about data touched upon in the section above, there were several objections both to how his theoretical framework explained the historical evidence and his predictions for likely trends in the future.

The most obvious point is that there is little evidence from Piketty’s data that r being greater than g leads to ever-increasing inequality. The gap between the rate of return on capital and the growth rate was larger in the

50

late 19th and early 20th centuries than now, or even than expected in the future, and though the level of inequality was higher at this time, it was fairly stable. So, clearly, the r > g aspect is not sufficient to explain the hypothesis that inequality will rise significantly in future.

In the mid-20th century, when the rate of return on capital was driven below the growth rate of the economy, Piketty asserts that this was due to the effects of wars, post-war policies and robust economic growth. But he does not calculate quantitatively the magnitudes of each of these factors – which obviously have very different policy implications. Can inequality be reduced by high tax rates and more government control of the economy, or is it necessary to subject the economy to deeply destructive conflicts? If the latter, then the means are far more destructive than the ends. Yet Piketty never really analyses the potential impact of his proposed policies on growth and prosperity more broadly – and we know that growth, rather than redistribution, has been the far more powerful mechanism for improving the condition of the poor over recent centuries.

In more recent history, it is widely acknowledged that the increase in total income inequality in the US in the past four decades has nothing to do with the returns to capital or the growth rate, but rather is reflective of rapid growth in labour income among the super-rich. Globalisation, technological change, the growth of finance and lower marginal tax rates on labour income may have all contributed to this (as did tax reforms which encouraged people to take more of their income as labour income – see Martin Feldstein). Little surprise then that among expert economists just two per cent agree with the statement ‘The most powerful force pushing towards greater wealth inequality in the US since the 1970s is the gap between the after-tax return on capital and the economic growth rate’, with 63 per cent disagreeing.

In fairness, Piketty’s book never claimed that the r > g phenomenon was the key factor responsible for rising inequality in recent times. His is a story more about the future. But many economists have been sceptical that his thesis will play out – not least because of widespread dissent against the idea of such a high elasticity of substitution between labour and capital. If instead of Piketty’s assumption that the elasticity of substitution is greater than 1.0 we assume it is less than 1.0 (as much of the literature suggests) then an increase in the capital to income ratio as we have seen in many countries can actually lead to a lower capital share of total income – actually lowering inequality using Piketty’s mechanism.

51

52

The r > g theory need not imply that inequality increases anyway. It depends on how much of capital income is consumed and the inheritance decisions of the rich. So r > g can be consistent with a stable wealth-to-income ratio, and stable inequality. In the UK, for example, there has been only a very small increase in wealth inequality since the 1970s despite the fact that the wealth-to-income ratio has increased significantly.

In fact, the overwhelming reason for the rise in the wealth-to-income ratio for countries such as the UK is rising housing wealth. But if rising housing wealth is the driver of the rising capital/income ratio then someone who finds themselves concerned about the impact on inequality might have much less to worry about.

The returns to housing capital are overwhelmingly consumed in the form of imputed rent and are not reinvested. Since housing in the UK is much more equally spread than other forms of wealth, we would expect this trend to have much less of an impact on inequality (at least to begin with – over time the inheritance transmission mechanism could change this). And we know a significant reason for high house prices in the UK is a lack of a supply response to changing demands, caused not least by extremely tight planning laws.

A rising capital-to-income ratio caused by rising housing wealth is therefore not primarily a consequence of capitalism. In fact, it is a failure of intervention and need not lead to rising inequality.

Over time, as these critiques have accumulated, the impact of Piketty’s work on the debate about inequality has diminished. We are now in a position where the majority of economists are highly sceptical of his thesis and framework for explaining trends in inequality – and in particular the strong claim that capitalism itself necessarily leads to rising inequalities. The accumulation of critical reviews about Piketty’s philosophical assumptions, data, theory and solutions mean that the debate about inequality is largely unchanged from what is what before: about whether it matters and what, if anything, can be done to combat it without adverse consequences for prosperity.

54

Institute ofEconomic A�airs

The Institute of Economic Affairs2 Lord North StreetLondon SW1P 3LBTel 020 7799 8900 email iea@iea.org.uk